Professional Documents
Culture Documents
Full Chapter Human Centred Intelligent Systems Proceedings of Kes Hcis 2020 Conference Alfred Zimmermann PDF
Full Chapter Human Centred Intelligent Systems Proceedings of Kes Hcis 2020 Conference Alfred Zimmermann PDF
https://textbookfull.com/product/intelligent-systems-and-
applications-proceedings-of-the-2020-intelligent-systems-
conference-intellisys-volume-2-kohei-arai/
https://textbookfull.com/product/intelligent-systems-and-
applications-proceedings-of-the-2020-intelligent-systems-
conference-intellisys-volume-3-kohei-arai/
https://textbookfull.com/product/proceedings-of-the-
international-conference-on-advanced-intelligent-systems-and-
informatics-2020-aboul-ella-hassanien/
https://textbookfull.com/product/intelligent-tutoring-
systems-16th-international-conference-its-2020-athens-greece-
june-8-12-2020-proceedings-vivekanandan-kumar/
Intelligent Human Systems Integration 2021 Proceedings
of the 4th International Conference on Intelligent
Human Systems Integration IHSI 2021 Integrating People
and Intelligent Systems February 22 24 2021 Palermo
Italy 1st Edition Dario Russo
https://textbookfull.com/product/intelligent-human-systems-
integration-2021-proceedings-of-the-4th-international-conference-
on-intelligent-human-systems-integration-ihsi-2021-integrating-
people-and-intelligent-systems-february-22-24/
https://textbookfull.com/product/electronic-systems-and-
intelligent-computing-proceedings-of-esic-2020-pradeep-kumar-
mallick/
https://textbookfull.com/product/intelligent-computing-
proceedings-of-the-2020-computing-conference-volume-3-kohei-arai/
Smart Innovation, Systems and Technologies 189
Alfred Zimmermann
Robert J. Howlett
Lakhmi C. Jain Editors
Human Centred
Intelligent
Systems
Proceedings of KES-HCIS 2020
Conference
123
Smart Innovation, Systems and Technologies
Volume 189
Series Editors
Robert J. Howlett, Bournemouth University and KES International,
Shoreham-by-sea, UK
Lakhmi C. Jain, Faculty of Engineering and Information Technology,
Centre for Artificial Intelligence, University of Technology Sydney,
Sydney, NSW, Australia
The Smart Innovation, Systems and Technologies book series encompasses the
topics of knowledge, intelligence, innovation and sustainability. The aim of the
series is to make available a platform for the publication of books on all aspects of
single and multi-disciplinary research on these themes in order to make the latest
results available in a readily-accessible form. Volumes on interdisciplinary research
combining two or more of these areas is particularly sought.
The series covers systems and paradigms that employ knowledge and intelligence
in a broad sense. Its scope is systems having embedded knowledge and intelligence,
which may be applied to the solution of world problems in industry, the environment
and the community. It also focusses on the knowledge-transfer methodologies and
innovation strategies employed to make this happen effectively. The combination of
intelligent systems tools and a broad range of applications introduces a need for a
synergy of disciplines from science, technology, business and the humanities. The
series will include conference proceedings, edited collections, monographs, hand-
books, reference books, and other relevant types of book in areas of science and
technology where smart systems and technologies can offer innovative solutions.
High quality content is an essential feature for all book proposals accepted for the
series. It is expected that editors of all accepted volumes will ensure that
contributions are subjected to an appropriate level of reviewing process and adhere
to KES quality principles.
** Indexing: The books of this series are submitted to ISI Proceedings,
EI-Compendex, SCOPUS, Google Scholar and Springerlink **
Lakhmi C. Jain
Editors
123
Editors
Alfred Zimmermann Robert J. Howlett
Faculty of Informatics KES International and Bournemouth
Reutlingen University University
Reutlingen, Baden-Württemberg, Germany Shoreham-by-sea, UK
Lakhmi C. Jain
University of Technology Sydney
Sydney, NSW, Australia
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore
Pte Ltd. 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Organisation
Honorary Chairs
General Chair
Executive Chair
Programme Chair
v
vi Organisation
Prof. Vincenzo Moscato, Università degli Studi di Napoli Federico II, Italy
Dr. Selmin Nurcan, University Paris Pantheon-Sorbonne, France
Prof. Andreas Oberweis, Karlsruhe Institute of Technology (KIT), Germany
Prof. Sofia Ouhbi, UAE University, UAE
Prof. Oscar Pastor Lopez, Universitat Politecnica de Valencia, Spain
Prof. Radu-Emil Precup, University of Timisoara, Romania
Prof. Carlos Ramos, ISEP/IPP, Portugal
Prof. Manfred Reichert, University of Ulm, Germany
Prof. Vaclav Repa, University of Economics, Czech Republic
Dr. Patrizia Ribino, National Research Council, Italy
Prof. Alexander Rossmann, Reutlingen University, Germany
Prof. Mohammed Sadgal, Cadi Ayyad University, Morocco
Prof. Kurt Sandkuhl, University of Rostock, Germany
Prof. Rainer Schmidt, Munich University of Applied Sciences, Germany
Dr. Christian Schweda, Reutlingen University, Germany
Prof. Sabrina Senatore, University of Salerno, Italy
Prof. Alberto Silva, University of Lisbon, Portugal
Dr. Stefano Silvestri, ICAR CNR, Italy
Dr. Milan Simic, RMIT University, Australia
Prof. Andreas Speck, University of Kiel, Germany
Dr. Maria Spichkova, RMIT University, Australia
Dr. Jim Spohrer, IBM Almaden Research, USA
Prof. Ulrike Steffens, Hamburg University of Applied Sciences, Germany
Prof. Janis Stirna, Stockholm University, Sweden
Prof. Eulalia Szmidt, Systems Research Institute Polish Academy of Sciences,
Poland
Prof. Hironori Takeuchi, Musahi University, Japan
Prof. Edmondo Trentin, University of Siena, Italy
Prof. Taketoshi Ushiama, Kyushu University, Japan
Prof. Rosa Vicari, Federal University of Rio Grande do Sul, Brazil
Prof. Toyohide Watanabe, Nagoya University, Japan
Dr. Jaroslaw Watrobski, University of Szczecin, Poland
Dr. Alicja Wieczorkowska, Polish-Japanese Academy of Information Technology,
Poland
Prof. Fons Wijnhoven, University of Twente, Netherlands
Prof. Matthias Wißotzki, Wismar University of Applied Sciences, Germany
Prof. Shuichiro Yamamoto, Nagoya University, Japan
Prof. Cecilia Zanni-Merk, INSA Normande University, France
Prof. Alfred Zimmermann, Reutlingen University, Germany
Preface
ix
x Preface
Editors
Reutlingen, Germany Alfred Zimmermann
Shoreham-by-sea, UK Robert J. Howlett
Sydney, Australia Lakhmi C. Jain
Contents
xi
xii Contents
xv
xvi About the Editors
Dr. Lakhmi C. Jain, Ph.D., M.E., Fellow (Engineers Australia) is affiliated with
the University of Technology Sydney, Australia, and Liverpool Hope University,
UK. Professor Jain serves with KES International, which provides the professional
community with opportunities for publication, knowledge exchange, cooperation,
and teambuilding. Involving over 5,000 researchers drawn from universities and
companies worldwide, KES facilitates international cooperation and generates
synergy in teaching and research. KES regularly provides networking opportunities
for the professional community through one of the largest conferences of its kind.
General Track of Human-Centred
Intelligent Systems or Human-Centred
Intelligent Systems
Visual Analytics Methods for Eye
Tracking Data
Abstract Nowadays, eye tracking data have become important and valuable infor-
mation that help to understand the behavior of users. The gathering of these data is not
an issue anymore. However, the problem is the analysis process and especially how
can these raw data be converted to understandable and useful information. Visual
analytics can solve this issue by combining human analytics skills and the advanced
computer analytics. This leads to the novel discoveries and helps humans take control
of the analytical process. These visualizations can be used to solve difficult problems
by discovering new unknown patterns of available data. In this work, we discussed
different methods that are used in the case of eye tracking data, and we addressed
the challenges of visual analytics in this context.
1 Introduction
Eye tracking systems are improving fast as the hardware advancement related to
these applications becomes more accessible for everybody at affordable prices. With
this, the gathered data from eye tracking devices trend to be considered big data. This
data is characterized by three parameters, such as volume, velocity, and variety. We
will describe in detail each parameter later in this paper.
N. Quadar
School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa,
ON K1N 6N5, Canada
e-mail: nquad016@uottawa.ca
A. Chehri (B)
Department of Applied Sciences, University of Québec in Chicoutimi, Chicoutimi,
QC G7H 2B1, Canada
e-mail: achehri@uqac.ca
G. Geon
Department Embedded Systems Engineering, Incheon National University, Incheon, Korea
e-mail: gjeon@inu.ac.kr
© The Editor(s) (if applicable) and The Author(s), under exclusive license 3
to Springer Nature Singapore Pte Ltd. 2021
A. Zimmermann et al. (eds.), Human Centred Intelligent Systems, Smart Innovation,
Systems and Technologies 189, https://doi.org/10.1007/978-981-15-5784-2_1
4 N. Quadar et al.
The challenge behind this evolution is how we can analyze this tremendous amount
of data. The statistical methods are limited to find an existing correlation between
different collected data. Also, traditional visualization methods such as scan path and
attention maps are restricted as the number of people being eye tracked increases.
Various studies have been conducted to investigate and test new approaches and
methods. Visual analytics is one of these alternative solutions to analyze and find
patterns between the enormous volumes of collected data. This method involves
different key components, such as analytic reasoning, visualization, and interaction
between computers and humans.
In this paper, we study the possibility of applying the visual analytics methods to
eye tracking data. Moreover, an example of possible future applications is provided,
and the last section discusses the future challenges of this movement.
Big data has a lot of different definitions, and one of them, according to Ward and
Barker is: “a term describing the storage and analysis of large and complex data
sets using series of techniques including, but not limited to NoSQL, MapReduce,
and machine learning” [1]. Another definition, according to De Mauro et al., is: “Big
Data represents the Information assets characterized by such a high volume, velocity,
and variety to require specific technology and analytic methods for its transformation
into value” [2].
According to these definitions and as mentioned previously, the volume, velocity,
and variety of the eye tracking data are increasing rapidly; therefore, they can be
considered as big data.
The Fig. 1 shows how big data and eye tracking data are connected. The eye
tracking data are becoming big data, and as visual analytics can be applied to big
data, why not use these methods to analyze this new kind of emerging data [3].
The common characteristics between big data and eye tracking data are their three
V’s, these parameters are increasing, and the collected data becomes more and more
complex and hard to analyze. A detailed explanation of these parameters is given
below.
Velocity: Different hardware and devices are now available to gather data from eye
movements. Smartphones, eye tracking glasses are some examples of this technology
that allow gathering data faster. Currently, the data movement in real time and in
a fraction of second, we can have a new update that should be implemented and
analyzed [4].
Volume: As the number of people having access to eye tracking systems increases and
with the recording speed of these new devices that know a considerable improvement
(e.g., SMI RED 500 having 500 Hz of recording rate) [3] the volume of data becomes
enormous. With crowdsourcing solutions and platforms, people can send their eye
tracking data from their homes and online. Also, the high-resolution cameras now
Visual Analytics Methods for Eye Tracking Data 5
that have been developed and continue to be improved will help to collect more data
in places where a lot of people are present such as soccer games.
Variety: Gathered data from eye tracking cannot be only the ones from eye move-
ment, but also other data could be collected, such as motion tracking, verbal data,
and so on.
These data help to understand the behavior of the tracker person and will complete
the information received from the eye movement. This may cause many problems
such as data synchronization, so new methods should be developed to take into
consideration all these various data in order to find the best patterns.
Nowadays, eye tracking is used in various applications that require to analyze the
user’s behavior, such as in marketing or human–computer interactions. Depends
on the application, a different type of collected data is generated by the recording
devices. The available tools can record with a rate up to 500 Hz [3] that depends on
the characteristic of each device.
The recording rate measures how many gaze points detected per second. Figure 2
shows different data types that can be gathered by an eye tracking system, such as
saccade, gaze, area of interest (AOIs), fixations, transition, and stimulus. Each type
is described below in detail.
Fixation: It is the combination of gaze points; two parameters define this combina-
tion: their combination area and timestamps. The timestamp or fixation duration is
6 N. Quadar et al.
the difference between the time when the eye to enter a specific point and the time
when the eye leaves this point (between 200 ms and 300 ms) [5]. The aggregation
area is usually around 20 and 50 pixels.
Saccade: The rapid movement of the eye when it jumps from the fixation on one
point to another is called the saccade; this movement can last up to 80 ms. Saccade
amplitude, duration, and velocity are typical metrics for this type of data.
Gaze: The sum of the fixation duration within a specific area.
Stimulus: It is the full region or area of study; in other words, it’s the visual content
where the eye movement will be tracked. It can be static or dynamic content in 2D
or 3D.
Areas of interest: Usually, the eye tracking applications are used to study the
behavior of people toward some specific regions. The areas of interest (AOIs) are
those regions in the stimulus that have high importance. The AOIs are defined based
on the stimulus semantics. The movement between different areas of interest is called
the transition. The AOIs can be seen, in the case of 3D stimulus as objects and are
called objects of interest (OOIs).
Data metrics and additional data sources: Using the different data types described
previously, various and more complex data metrics can be generated, such as direc-
tions of the saccades, fixations numbers, saccade amplitude, and so on. For instance,
analyzing the time variation of some data metrics such as fixation duration over time
can lead to understanding the learning effect while checking a stimulus. The addi-
tional data source can be used to understand the tracked person’s behavior; these
data can be the mouse movement or, moreover, the interaction with the keyboard in
the case of using a computer to show the stimulus. The social media interaction can
Visual Analytics Methods for Eye Tracking Data 7
also be used in the case of smartphone use. These data are synchronized with other
data collected from the eye tracking devices and can be useful to find a connection
between tracked people.
This section discusses different methods based on visual analytics science and
explains different steps before and during the data analysis. As mentioned before,
there are various data collected from the eye tracking devices. The sequence order
over the time of different data is called the trajectory or the scan path; it can also
be defined on how the movement changes over time when the person is under-eye
tracking. The following method will be described [6]:
• Heat map
• Space–time cube
• Gaze plot.
Before analyzing eye tracking data, various transformation should be done in order
to make the analysis easier and efficient [7]. One of these transformations is the
adjustment of the time reference: it aims, in the context of eye tracking data analysis,
to align the start or the end times of multiples trajectories by shifting the trajectories
timeline to the same end or start without changing any variants.
The second transformation to perform is the spatial generalization, which aims to
replace the trajectories’ spatial positions by new ones having the same space units
(points by area). The last transformation is the spatiotemporal aggregation; in this
step, the paths will be transformed into a series of visited places and moves between
these places. For this purpose, various statistic parameters are computed such as the
count of visits and users, average time spent per place, and so on.
4.2 Methods
Visual analysis methods are used to fill the gap in statistical methods and add addi-
tional insights. Visualization is always a good tool to use in the case of spatiotemporal
data.
There are two, as mentioned above, techniques used in visualization analysis: the
heat map, space–time cube, and gaze plot.
8 N. Quadar et al.
• Heat map:
The attention map or the heat map provides an overview of the aggregation of gaze
points within the stimulus in order to reveal the distribution of visual attention.
Therefore, the heat map helps to identify the main areas of interest.
As we can see from Fig. 3, the attention map outputs are a color-coded scheme.
Red areas show a high number of gaze points, which means essential areas of interest,
green area have less gaze points. The results are obtained by taking the aggregation of
gaze position over the observation time. However, this method has some limitations,
such as it doesn’t take into consideration the temporal data component; also, it has a
weak efficiency for dynamic images [8].
• Space–time cube:
• Gaze plot:
The gaze plot, also known as scan path, uses the position of the fixation points and
fixation time to produce an overview of the sequence of user’ fixations. This method
represents the fixation by circles with size proportional to the fixation duration, as
can be seen from Fig. 5. Moreover, it connects fixation circles bylines to give the
exact sequence of the trajectories.
The eye tracking data can help the supermarket to understand their buying behavior
and strategy by analyzing their eye movement and finding patterns between different
data using the methods described above. The tracking can be done by using high-
definition cameras or by giving eye tracking glasses to customers at the entrance of the
supermarket. Using devices such as eye tracking glasses can be used to analyze real-
time data and take live decisions such as guide users to production on sale or giving
suggestions on what to buy. Using a high-definition camera would be challenging as
the stimulus will be modeled on 3D, and different users will have a different stimulus.
(2) Driving case study
As the number of cars and the accident related to driving increases, a new solution
such as eye tracking is necessary to face this challenge. In the future, cars can be
smarter by integrating the eye tracking system to their control system. This will lead
to analyze the traffic situation and allow the car to act depending on the situation.
With such a solution, driving can be safer and more comfortable.
6 Future Challenges
As we discussed in the previous sections, the amount of eye tracking data is increasing
day after day, and the number of people having access to this technology increases
as well. Combining and synchronizing these data present a big challenge, especially
when the stimulus content is dynamic and changing with time [9]. Below a summary
of the main problems that will face this technology in the future:
Stimuli: The content is becoming more and more 3D, and the stimuli can be
different from user to user. So more advanced technologies are needed to satisfy
all requirements.
Users: As mentioned, the number of users is increasing, and more data will be
generated. So, new procedures should be set up as the users will be non-experts and
will need more reliable methods to analyze their data.
Hardware: New hardware devices are getting into the market and keep improving,
so the analysis techniques should take into consideration different aspect of the newly
available equipment [10–26].
Privacy: One of the coming and significant issues is the people’s privacy as the
number of users increases. The new mechanism should be applied to protect the
millions of personal data that should not be available to anyone.
Table 1 shows an evolution of various challenges over time.
Visual Analytics Methods for Eye Tracking Data 11
7 Conclusion
In this paper, we have discussed the trend of the eye tracking data and how they are
considered big data. We described a different type of data that can be generated using
this technology and its principal characteristic parameters, such as velocity, variety,
and volume. Moreover, we discussed the different visual analytics methods that can
be applied in different scenarios.
Based on our references, this field still has many challenges to face, such as the
need for new analysis techniques that can be used in the case of the variety of the
data collected and their huge volume.
References
1. Ward, J.S., Barker, A.: Undefined by data: a survey of big data definitions. CoRR,
abs/1309.5821, pp. 1–2 (2013)
2. Hashem I.A.T., Yaqoob, I., Anuar, N.B., Mokhtar, S., Gani, A., Khan, S.U.: The rise of ‘big
data’ on cloud computing: review and open research issues. Inform. Syst. 47, 98–115 (2015)
3. Blascheck, T., Burch, M., Raschke, M., Weiskopf, D.: Challenges and perspectives in big
eye-movement data visual analytics. In: Big Data Visual Analytics (BDVA), pp. 1–8, 22–25
(2015)
4. De Mauro, A., Greco, M., Grimaldi, M.: What is big data? A consensual definition and a review
of key research topics. In: AIP Conference Proceedings, pp. 97–104. AIP Publishing (2015)
5. Blascheck, T., Kurzhals, K., Raschke, M., Burch, M., Weiskopf, D., Ertl, T.: State-of-the-art
of visualization for eye tracking data. In: Borgo, R., Maciejewski, R., Viola, I. (eds.) EuroVis–
STARs, pp. 63–82 (2014)
6. Andrienko, G.L., Andrienko, N.V., Burch, M., Weiskopf, D.: Visual analytics methodology for
eye movement studies. IEEE Trans. Vis. Comput. Graph. 18(12), 2889–2898 (2012)
7. Andrienko, G., Andrienko, N., Bak, P., Keim, D., Kisilevich, S., Wrobel, S.: A conceptual
framework and taxonomy of techniques for analyzing movement. J. Vis. Lang. Comput. 22(3),
213–223 (2011)
12 N. Quadar et al.
8. Kurzhals, K., Weiskopf, D.: Space-time visual analytics of eye-tracking data for dynamic
stimuli. IEEE Trans. Vis. Comput. Graph. 19(12), 2129–2138 (2013)
9. Stellmach, S., Nacke, L., Dachselt, R.: Advanced gaze visualizations for three-dimensional
virtual environments. In: Proceedings of the Symposium on Eye Tracking Research &
Applications, pp. 109–112 (2010)
10. Kim, Y., Varshney, A.: Persuading visual attention through geometry. IEEE Trans. Vis. Comput.
Graph. 14(4), 772–782 (2008)
11. Klingner, J., Kumar, R., Hanrahan, P.: Measuring the task-evoked pupillary response with a
remote eye tracker. In: Proceedings of the Symposium on ETRA, pp. 69–72 (2008)
12. Kurzhals, K., Heimerl, F., Weiskopf, D.: ISeeCube: visual analysis of gaze data for video. In:
Proceedings of the Symposium on ETRA, pp. 43–50 (2014)
13. Kurzhals, K., Hoferlin, M., Weiskopf, D.: Evaluation of attention-guiding video visualization.
Comput. Graph. Forum 32(3), 51–60 (2013)
14. Lam, H., Bertini, E., Isenberg, P., Plaisant, C., Carpendale, S.: Empirical studies in information
visualization: seven scenarios. IEEE Trans. Vis. Comput. Graph. 18(9), 1520–1536 (2012)
15. Liu, G., Austen, E.L., Booth, K.S., Fisher, B.D., Argue, R., Rempel, M., Enns, J.T.: Multiple-
object tracking is based on scene, not retinal, coordinates. J. Exp. Psychol. Hum. Percept.
Perform. 31(2), 235–247 (2005)
16. Loftus, G.R., Mackworth, N.H.: Cognitive determinants of fixation location during picture
viewing. J. Exp. Psychol. Hum. Percept. Perform. 4(4), 565–572 (1978)
17. Milner, A.D., Goodale, M.A.: Two visual systems reviewed. Neuropsychologia 46(3), 774–785
(2008)
18. Plaisant, C.: The challenge of information visualization evaluation. In: Proceedings of the
Working Conference on Advanced Visual Interfaces, pp. 109–116 (2004)
19. Po, B.A., Fisher, B.D., Booth, K.S.: Pointing and visual feedback for spatial interaction in large-
screen display environments. In: Proceedings of the Third International Symposium Smart Grid.
Lecture Notes in Computer Science, pp. 22–38. Springer (2003)
20. Poole, A., Ball, L.: Eye tracking in human-computer interaction and usability research: current
status and future prospects. In: Ghaoui, C. (ed.) Encyclopedia of Human-Computer Interaction,
pp. 211–219. Idea Group Inc. (2006)
21. Ribarsky, W., Fisher, B.D., Pottenger, W.M.: Science of analytical reasoning. Infom. Vis. 8(4),
254–262 (2009)
22. Ristovski, G., Hunter, M., Olk, B., Linsen, L.: EyeC: coordinated views for interactive
visual exploration of eye-tracking data. In: Proceedings of the Conference on Information
Visualization (IV), pp. 239–248 (2013)
23. Song, H., Yun, J., Kim, B., Seo, J.: GazeVis: interactive 3D gaze visualization for contiguous
cross-sectional medical images. IEEE Trans. Vis. Comput. Graph. 20(5), 726–739 (2014)
24. Spence, B.: The broker. In: Ebert, A., Dix, A., Gershon, N.D., Pohl, M. (eds.) Human Aspects
of Visualization. Lecture Notes in Computer Science, vol. 6431, pp. 10–22. Springer (2011)
25. Star, S.L.: The structure of ill-structured solutions: boundary objects and heterogeneous
distributed problem solving. In: Huhns, M., Gasser, L. (eds.) Readings in Distributed AI.
Morgan Kaufmann (1988)
26. Swindells, C., Tory, M., Dreezer, R.: Comparing parameter manipulation with mouse, pen, and
slider user interfaces. Comput. Graph. Forum 28(3), 919–926 (2009)
27. Hurter, C., Ersoy, O., Fabrikant, S., Klein, T., Telea, A.: Bundled visualization of dynamic
graph and trail data. IEEE Trans. Vis. Comput. Graph. (2013)
28. Jarodzka, H., Holmqvist, K., Nyström, M.: A vector-based, multidimensional scanpath simi-
larity measure. In: Proceedings of the 2010 Symposium on Eye-Tracking Research &
Applications, pp. 211–218 (2010)
Challenges of Adopting Human-Centered
Intelligent Systems: An Organizational
Learning Approach
Fons Wijnhoven
1 Introduction
Clinical decision support systems (CDSSs) are systems for clinical decision making
[1]. A CDSS can contain multiple techniques to support medical decision-making,
like visualization of clinical data [2, 3], zooming, sorting and filtering to deep dive
in specific sections of relevant patient data, images and video recordings [4–6] and
analysis of data from multiple sources like the patient medical record and genomic
data for diagnosis and treatment selection [7]. Natural language processing (NLP)
can also be used to extract the meaning from natural language text notes in medical
records [8–10]. More complex forms of technology involve machine learning, i.e.
methods that can automatically detect patterns in data [11]. CDSSs like Watson for
Oncology incorporate a form of prescriptive analytics by ranking treatment alterna-
tives along predicted effectiveness for a given diagnosis with the support of mining
knowledge from 600 medical journals, hundreds of different medical data sources
F. Wijnhoven (B)
University of Twente, Enschede, The Netherlands
e-mail: a.b.j.m.wijnhoven@utwente.nl
© The Editor(s) (if applicable) and The Author(s), under exclusive license 13
to Springer Nature Singapore Pte Ltd. 2021
A. Zimmermann et al. (eds.), Human Centred Intelligent Systems, Smart Innovation,
Systems and Technologies 189, https://doi.org/10.1007/978-981-15-5784-2_2
14 F. Wijnhoven
Table 1 Analytics
Technique Healthcare application examples
techniques in clinical DSS
[13] Cluster analysis Detecting high-risk obesity groups
Machine learning Predicting disease risk; Detecting
epidemics
Neural networks Diagnosing chronic diseases; Prediction
patients’ future diseases
Pattern recognition Improving public health surveillance
and statistical evidence [12]. Mehta and Pandit summarize analytic techniques for
CDSS in Table 1 [13].
Mehta and Pandit [13] state that all the studies they found describe, develop
and test a model or algorithm to show its added value, however, do not mention
anything about implementation. Mehta and Pandit [13] suggest that a reason for
the lack of these CDSS implementations is that the current body of literature does
not provide adequate quantitative evidence that these techniques can be trusted by
medical practitioners in their clinical use. Also other review articles state that different
from research contexts, medical clinical contexts have very high levels of ethical,
legal and reasoning transparency demands that are difficult to meet in practice [14].
A lack of trust in the systems recommendations is also an important reason for not
adopting CDSS by physicians [14]. Highly problematic is also medical journalists
claim that Watson for Oncology’s training resulted in a bias towards Memorial Sloan
Kettering physicians’ preferences, which is not unlikely because these physicians
were involved in the development of Watson for Oncology [15, 16]. Other technology
companies such as Google and Microsoft have developed similar CDSS [17]. These
systems remain within the R&D environment and have not been implemented in
clinical practice either [18].
Despite these adoption problems, McNut et al. [19] and Rumsfeld et al. [20] state
that analytics has great promises for the fields of oncology and cardiovascular diag-
noses and treatments, respectively, but that these promises only can become realized
when sufficiently large and reliable datasets are available. Unfortunately, this is diffi-
cult to achieve. For realizing such large data sets hospitals will have to share their data
[21–23]. Lack of systems interoperability and different data taxonomies prevent such
inter-hospital data sharing and thus further structuring and standardization of systems
and data is needed [7, 24]. NLP of informal language that describes patient’s status
and medical doctor’s thoughts are a new opportunity for CDSS. These data are avail-
able via patient medical records, but accessibility of these data for analytics is legally
and ethically complicated and classifying natural language in medical terminology
is still not very reliable [25]. Besides of all these analytical workflow challenges,
CDSSs are also difficult to use from the perspective of the medical practitioner.
Medical practitioners are especially concerned regarding the accuracy of classifica-
tions and predictions when the data sets are too small [26–30] and the algorithms
used are non-transparent or incomprehensible [31]. Intransparency of algorithms
may result in feelings of loss of reasoning control which is unacceptable for medical
Another random document with
no related content on Scribd:
Fig. 241.—Lips of a Wrasse,
Labrus festivus.
The “Wrasses” are a large family of littoral fishes, very abundant
in the temperate and tropical zones, but becoming scarcer towards
the Arctic and Antarctic circles, where they disappear entirely. Many
of them are readily recognised by their thick lips, which are
sometimes internally folded, a peculiarity which has given to them
the German term of “Lip-fishes.” They feed chiefly on mollusks and
crustaceans, their dentition being admirably adapted for crushing
hard substances. Many species have a strong curved tooth at the
posterior extremity of the intermaxillary, for the purpose of pressing a
shell against the lateral and front teeth by which it is crushed. Other
Wrasses feed on corals, others on zoophytes; a few are herbivorous.
In all Wrasses the upper pharyngeal bones seem to be jointed to the
basi-occipital; but whilst in Labrus the basi-occipital is raised on each
side into a large flattish condyle, fitting into a concavity of the upper
pharyngeals, in Scarus the mode of articulation is reversed, the basi-
occipital having a pair of long grooves, in which the oblong condyles
of the upper pharyngeals slide forwards and backwards. Beautiful
colours prevail in this family, permanent pigmentary colours as well
as passing iridescent reflections of the scales. Some species remain
very small, others grow to a weight of fifty pounds. The larger kinds
especially are prized as food, the smaller less so.
Remains of Labridæ, recognised by their united pharyngeals,
which bear molar-like teeth, are not scarce in tertiary formations of
France, Germany, Italy, and England. Such remains from Monte
Bolca and the Swiss Molasse have been referred to the genus
Labrus. Others, Nummopalatus and Phyllodus, are allied, but cannot
be assigned, to one of the recent genera; the latter genus is first
represented in cretaceous formations of Germany. Another genus,
Taurinichthys, from the Miocene of France, represents the Odacina
of the living fauna. Egertonia, from the Isle of Sheppey, differs so
much from all recent Labroid genera that its pertinence to this family
appears doubtful.
[See J. Cocchi, Monografia dei Pharyngodopilidæ, 1866; and E.
Sauvage, Sur le genre Nummopalatus, in Bull. Soc. Geol. France,
1875.]
Labrus.—Body compressed, oblong, covered with scales of
moderate size, in more than forty transverse series; snout more or
less pointed; imbricate scales on the cheeks and opercles; none or
only a few on the interoperculum. Teeth in the jaws conical, in a single
series. Dorsal spines numerous, thirteen or twenty-one, none of which
are prolonged; anal spines three. Lateral line not interrupted.
Young “Wrasses” differ from mature specimens in having the
præoperculum serrated. The headquarters of this genus are the
Mediterranean, whence it ranges, gradually diminishing towards the
north, along all the shores of Europe. Nine species are known;
British are the “Ballan Wrasse” (L. maculatus), and the “Striped or
Red Wrasse” or “Cook” (L. mixtus). The two sexes of the latter
species are very differently coloured; the male being generally
ornamented with blue streaks, or a blackish band along the body,
whilst the female has two or three large black blotches across the
back of the tail.
Crenilabrus are Labrus with serrated præoperculum; the number
of their dorsal spines varies from thirteen to eighteen, and the scales
are arranged in less than forty transverse series.
The range of this genus is co-extensive with Labrus. C. melops,
the “Gold-sinny,” or “Cork-wing,” is common on the British coasts.
Tautoga.—Body compressed, oblong, covered with small scales;
scales on the cheek rudimentary, opercles naked. Teeth in the jaws
conical, in double series; no posterior canine tooth. Dorsal spines
seventeen, anal spines three. Lateral line not interrupted.
The “Tautog,” or “Black-fish,” is common on the Atlantic coasts of
temperate North America, and much esteemed as food.
Ctenolabrus.—Body oblong, covered with scales of moderate
size; imbricate scales on the cheeks and opercles. Teeth in the jaws in
a band, with an outer series of stronger conical teeth; no posterior
canine tooth. Dorsal spines from sixteen to eighteen; anal spines
three. Lateral line not interrupted.
Four species, from the Mediterranean and the temperate parts of
the North Atlantic, Ct. rupestris being common on the British, and Ct.
burgall on the North American coasts.
Acantholabrus.—A Wrasse with five or six anal spines, and with
the teeth in a band.
Two species are known from Madeira and the Canary Islands,
and one from northern Europe and Greenland. The latter is scarce
on the British coasts, but bears a distinct name on the south coast,
where it is called “Rock-cook.”
Lachnolaemus from the West Indies, and Malacopterus from
Juan Fernandez, are Labroids, closely allied to the preceding North
Atlantic genera.
Cossyphus.—Body compressed, oblong, with scales of moderate
size; snout more or less pointed; imbricate scales on the cheeks and
opercles; basal portion of the vertical fins scaly. Lateral line not
interrupted. Teeth in the jaws in a single series; four canine teeth in
each jaw anteriorly; a posterior canine tooth. Formula of the fins: D.
12/9–11, A. 3/12.
Twenty species are known from the tropical zone and coasts
adjoining it; some, like G. gouldii from Tasmania, attain a length of
three or four feet.
Chilinus.—Body compressed, oblong, covered with large scales;
lateral line interrupted; cheeks with two series of scales;
præoperculum entire; teeth in a single series, two canines in each jaw;
no posterior canine tooth; lower jaw not produced backwards. Dorsal
spines subequal in length; formula of the fins: D. 9–10/10–9, A. 3/8.
Common in the tropical Indo-Pacific, whence more than twenty
species are known. Hybrids between the different species of this
genus are not uncommon.
Epibulus.—Closely allied to the preceding genus, but with a very
protractile mouth, the ascending branches of the intermaxillaries, the
mandibles, and the tympanic being much prolonged.
This fish (E. insidiator) is said to seize marine animals by
suddenly thrusting out its mouth and engulphing those that come
within the reach of the elongated tube. It attains a length of twelve
inches, is common in the tropical Indo-Pacific, and varies much in
coloration.
Anampses.—Distinguished by its singular dentition, the two front
teeth of each jaw being prominent, directed forwards, compressed,
with cutting edge. D. 9/12, A. 3/12.
Beautifully coloured fishes from the tropical Indo-Pacific. Ten
species.
Platyglossus.—Scales in thirty or less transverse series; lateral
line not interrupted. A posterior canine tooth. Dorsal spines nine.
Small beautifully coloured Coral-fishes, abundant in the
equatorial zone and the coasts adjoining it. Some eighty species are
known (inclusive of the allied genera Stethojulis, Leptojulis, and
Pseudojulis).
Novacula.—Body strongly compressed, oblong, covered with
scales of moderate size; head compressed, elevated, obtuse, with the
supero-anterior profile more or less parabolic; head nearly entirely
naked. Lateral line interrupted. No posterior canine tooth. D. 9/12, A.
3/12; the two anterior dorsal spines sometimes remote or separate
from the others.
Twenty-six species are known from the tropical zone, and the
warmer parts of the temperate zones. They are readily recognised
by their compressed, knife-shaped body, and peculiar physiognomy;
they scarcely exceed a length of twelve inches.
Julis.—Scales of moderate size; lateral line not interrupted. Head
entirely naked. Snout of moderate extent, not produced; no posterior
canine tooth. Dorsal spines ten.
Co-extensive with Platyglossus in their geographical distribution,
and of like beautiful coloration and similar habits. Some of the most
common fishes of the Indo-Pacific, as J. lunaris, trilobata, and
dorsalis, belong to this genus.
Coris.—Scales small, in fifty or more transverse series; lateral line
not interrupted. Head entirely naked. Dorsal spines nine.
Twenty-three species, distributed like Platyglossus; two reach the
south coast of England, Coris julis and C. giofredi, said to be male
and female of the same species. Some belong to the most
gorgeously coloured kinds of the whole class of fishes.
Genera allied to the preceding Labroids are—Choerops,
Xiphochilus, Semicossyphus, Trochocopus, Decodon, Pteragogus,
Clepticus, Labrichthys, Labroides, Duymæria, Cirrhilabrus,
Doratonotus, Pseudochilinus, Hemigymnus, Gomphosus, Cheilio,
and Cymolutes.
Pseudodax.—Scales of moderate size; lateral line continuous;
cheeks and opercles scaly. Each jaw armed with two pairs of broad
incisors, and with a cutting lateral edge; teeth of the lower pharyngeal
confluent, pavement-like. Dorsal spines eleven.
The fishes of this genus, and the three succeeding, are known by
the name of “Parrot-wrasses.” Of Scarus one species (S. cretensis)
occurs in the Mediterranean, and nine others in the tropical Atlantic.
The first was held in high repute by the ancients, and Aristotle has
several passages respecting its rumination. It was most plentiful and
of the best quality in the Carpathian Sea, between Crete and Asia
Minor, but was not unknown even in early times on the Italian coasts,
though Columella says that it seldom passed beyond Sicily in his
day. But in the reign of Claudius, according to Pliny, Optatus
Elipentius brought it from the Troad, and introduced it into the sea
between Ostium and Campagna. For five years all that were caught
in the nets were thrown into the sea again, and from that time it was
an abundant fish in that locality. In the time of Pliny it was considered
to be the first of fishes (Nunc Scaro datur principatus); and the
expense incurred by Elipentius was justified, in the opinion of the
Roman gourmands, by the extreme delicacy of the fish. It was a fish,
said the poets, whose very excrements the gods themselves were
unwilling to reject. Its flesh was tender, agreeable, sweet, easy of
digestion, and quickly assimilated; yet if it happened to have eaten
an Aplysia, it produced violent diarrhœa. In short, there is no fish of
which so much has been said by ancient writers. In the present day
the Scarus of the Archipelago is considered to be a fish of exquisite
flavour; and the Greeks still name it Scaro, and eat it with a sauce
made of its liver and intestines. It feeds on fucus; and Valenciennes
thinks that the necessity for masticating its vegetable diet thoroughly,
and the working of it with that intent backwards and forwards in the
mouth, may have given rise to the notion of its being a ruminant; and
it is certain that its aliment is very finely divided when it reaches the
stomach.
Fig. 242.—Scarichthys auritus.
Scarichthys.—Differing from Scarus only in having flexible dorsal
spines.
Third Family—Embiotocidæ.
Body compressed, elevated or oblong, covered with cycloid
scales; lateral line continuous. One dorsal fin, with a spinous portion,
and with a scaly sheath along the base, which is separated by a
groove from the other scales; anal with three spines and numerous
rays; ventral fins thoracic, with one spine and five rays. Small teeth
in the jaws, none on the palate. Pseudobranchiæ present. Stomach
siphonal, pyloric appendages none. Viviparous.
Marine Fishes characteristic of the fauna of the temperate North
Pacific, the majority living on the American side, and only a few on
the Asiatic. All are viviparous (see Fig. 70, p. 159). Agassiz
describes the development of the embryoes as a normal ovarian
gestation, the sac containing the young not being the oviduct but the
ovarian sheath, which fulfils the functions of the ovary. This organ
presents two modes of arrangement: in one there is a series of
triangular membranous flaps communicating with each other,
between which the young are arranged, mostly longitudinally, the
head of one to the tail of another, but sometimes with the bodies
curved, to the number of eighteen or twenty; in the other, the cavity
is divided by three membranes converging to a point, into four
compartments, not communicating with each other except towards
the genital opening, the young being arranged in the same
longitudinal manner. The proportionate size of the young is very
remarkable. In a female specimen 10½ inches long, and 4½ inches
high, the young were nearly 3 inches long and 1 inch high.
Seventeen species are known, the majority of which belong to
Ditrema, and one to Hysterocarpus. They do not attain to a large
size, varying from three-quarters to three pounds in weight.
Fourth Family—Chromides.
Body elevated, oblong or elongate, scaly, the scales being
generally ctenoid. Lateral line interrupted or nearly so. One dorsal
fin, with a spinous portion; three or more anal spines; the soft anal
similar to the soft dorsal. Ventral fins thoracic, with one spine and
five rays. Teeth in the jaws small, palate smooth. Pseudobranchiæ
none. Stomach coecal; pyloric appendages none.
Freshwater-fishes of rather small size from the tropical parts of
Africa and America; one genus from Western India. The species with
lobate teeth, and with many circumvolutions of the intestines, are
herbivorous, the other carnivorous.
Etroplus.—Body compressed, elevated, covered with ctenoid
scales of moderate size. Lateral line indistinct. Dorsal and anal spines
numerous. Teeth compressed, lobate, in one or two series. Anterior
prominences of the branchial arches not numerous, short, conical,
hard. Dorsal fin not scaly.
Two species from Ceylon and Southern India.
Chromis.—Body compressed, oblong, covered with cycloid scales
of moderate size. Dorsal spines numerous, anal spines three. Teeth
compressed, more or less lobate, in one series. Anterior prominences
of the branchial arches short, thin, lamelliform, non-serrated. Dorsal
fin not scaly.
Some twenty species are known from the fresh waters of Africa
and Palestine; the most celebrated is the “Bulti,” or “Bolty,” of the
Nile, one of the few well-flavoured fishes of that river; it grows to the
length of twenty inches. Two or three species of this genus occur in
the Jordan and Lake of Galilee.
Fig. 244.—Chromis andreæ, from the Lake of Galilee.
Hemichromis, differing from Chromis in having conical teeth in
one or two series.
THIRD ORDER—ANACANTHINI.
Vertical and ventral fins without spinous rays. The ventral fins, if
present, are jugular or thoracic. Air-bladder, if present, without
pneumatic duct.
These characters are common to all the members of this order,
with the exception of a freshwater-fish from Tasmania and South
Australia (Gadopsis), which has the anterior portion of the dorsal and
anal fins formed of spines.
First Family—Lycodidæ.
Vertical fins confluent. Ventral fin, if present, small, attached to
the humeral arch, jugular. Gill-opening narrow, the gill-membrane
being attached to the isthmus.
Marine littoral fishes of small size, resembling Blennies, chiefly
represented in high latitudes, but a few living within the tropical zone.
Lycodes.—Body elongate, covered with minute scales imbedded
in the skin, or naked; lateral line more or less indistinct. Eye of
moderate size. Ventral small, short, rudimentary, jugular, composed of
several rays. Upper jaw overlapping the lower. Conical teeth in the
jaws, on the vomer, and on the palatine bones. Barbel none. Five or
six branchiostegals; gill-opening narrow, the gill-membranes being
attached to the isthmus. Pseudobranchiæ present. Air-bladder none.
Pyloric appendages two, or rudimentary, or entirely absent. No
prominent anal papilla.
Fig. 246.—Lycodes mucosus, from Northumberland Sound.
Nine species are known from the Arctic Ocean, four from the
southern extremity of the American continent.
Gymnelis.—Body elongate, naked. Eye of moderate size or rather
small. Ventrals none. Vent situated at some distance backwards from
the head. Small conical teeth in the jaws, on the vomer and palatine
bones. Jaws equal anteriorly. Barbel none. Six branchiostegals; gill-
opening narrow, the gill membranes being attached to the isthmus.
Pseudobranchiæ present; air-bladder none. Pyloric appendages two;
no prominent anal papilla.
Second Family—Gadidæ.
Body more or less elongate, covered with small smooth scales.
One, two, or three dorsal fins, occupying nearly the whole of the
back; rays of the posterior dorsal well developed; one or two anal
fins. Caudal free from dorsal and anal, or, if they are united, the
dorsal with a separate anterior portion. Ventrals jugular, composed of
several rays, or, if they are reduced to a filament, the dorsal is
divided into two. Gill-opening wide; the gill-membranes generally not
attached to the isthmus. Pseudobranchiæ none, or glandular,
rudimentary. An air-bladder and pyloric appendages generally
present.
The family of “Cod-fishes” consists partly of littoral and surface
species (and they form the majority), partly of deep-sea forms. The
former are almost entirely confined to the temperate zones,
extending beyond the Arctic Circle; the latter have, as deep-sea
fishes generally, a much wider range, and hitherto have been found
chiefly at considerable depths of rather low latitudes. Only two or
three species inhabit fresh waters. They form one of the most
important articles of food and subsistence to the fishermen in Europe
and North America, and to whole tribes bordering upon the Arctic
Ocean.
Fossil remains are scarce. Nemopteryx and Palæogadus have
been described from the schists of Glaris, a formation believed to
have been the bottom of a very deep sea. In the clay of Sheppey
species occur allied to Gadus, Merluccius, and Phycis; others, not
readily determinable, have been found at Licata in Sicily (Miocene).
Gadus.—Body moderately elongate, covered with small scales. A
separate caudal, three dorsal, and two anal fins; ventrals narrow,
composed of six or more rays. Teeth in the upper jaw in a narrow
band; vomerine teeth; none on the palatines.
Arctic and temperate zones of the Northern Hemisphere.
Eighteen species are known, of which the following are the most
important:—
Gadus morrhua, the common “Cod-fish”—in German called
“Kabeljau” when fresh and old, “Dorsch” when young and fresh,
“Stock-fish” when dried, “Labberdan” when salted—measures from
two to four feet, and attains to a weight of one hundred pounds. On
the British coasts and in the German Ocean it is generally of a
greenish or brownish-olive colour, with numerous yellowish or brown
spots. Farther northwards darker-coloured specimens, frequently
without any spots, predominate; and on the Greenland, Iceland, and
North Scandinavian coasts the Cod have often a large irregular black
blotch on the side. The Cod-fish occurs between 50° and 75° lat. N.,
in great profusion, to a depth of 120 fathoms, but is not found nearer
the Equator than 40° lat. Close to the coast it is met with singly all
the year round, but towards the spawning-time it approaches the
shore in numbers, which happens in January in England and not
before May on the American coasts. The English resorted to the cod-
fisheries of Iceland before the year 1415, but since the sixteenth
century most vessels go to the banks of Newfoundland, and almost
all the preserved Cod consumed during Lent in the various
continental countries is imported from across the Atlantic. At one
time the Newfoundland cod-fishery rivalled in importance the whale-
fishery and the fur trade of North America. Cod-liver oil is prepared
from the liver on the Norwegian coast, but also other species of this
genus contribute to this most important drug.
Gadus tomcodus abundantly occurs on the American coasts; it
remains within smaller dimensions than the common Cod-fish.
Gadus æglefinus, the “Haddock” (“Schell-fisch” of the Germans,
“Hadot” of the French), is distinguished by a black lateral line and a
blackish spot above the pectoral fin. It attains to a length of three feet
in the higher latitudes, but remains smaller on more southern coasts;
like the Cod it extends across the Atlantic. The largest specimens
are taken on the British coast in winter, because at that time they
leave the deep water to spawn on the coast. Gadus merlangus, the
“Whiting,” with a black spot in the axil of the pectoral fin. Gadus
luscus, the “Bib,” “Pout,” or “Whiting-pout,” with cross-bands during
life, and with a black axillary spot, rarely exceeding a weight of five
pounds. Gadus fabricii, a small species, but occurring in incredible
numbers on the shores near the Arctic circle, and ranging to 80° lat.
N. Gadus pollachius, the “Pollack,” without a barbel at the chin, and
with the lower jaw projecting beyond the upper. Gadus virens, the
“Coal-fish,” valuable on account of its size and abundance, and
therefore preserved for export like the Cod.
The fishes of the genus Gadus are bathymetrically succeeded by
several genera, as Gadiculus, Mora, and Strinsia; however these do
not descend to sufficiently great depths to be included into the deep-
sea Fauna; the two following are true deep-sea fishes.
Halargyreus.—Body elongate, covered with small scales. Two
dorsal and two anal fins; ventrals composed of several rays. Jaws with
a band of minute villiform teeth; vomer and palatines toothless. No
barbel.
The single species known, H. johnsonii, proves to be a deep-sea
fish by its organisation as well as geographical distribution. Originally
known from a single specimen, which was obtained at Madeira, it
has since been found off the coast of New Zealand. There is no
doubt that it will be discovered also in intermediate seas.
Melanonus.—Head and body rather compressed, covered with
cycloid scales of moderate size, and terminating in a long tapering tail,
without caudal. Eye of moderate size. Villiform teeth in the jaws, on
the vomer and palatine bones. Barbel none. A short anterior dorsal,
the second extending to the end of the tail, and the anal being of
similar length. Ventrals composed of several rays. Bones soft and
flexible.
This is one of the discoveries made during the expedition of the
“Challenger.” The single specimen obtained is of a deep-black
colour, and was dredged up at a depth of 1975 fathoms in the
Antarctic Ocean.
Merluccius.—Body elongate, covered with very small scales. A
separate caudal; two dorsal fins and one anal; ventrals well
developed, composed of seven rays. Teeth in the jaws and on the
vomer rather strong, in double or triple series. No barbel.
Two species are known of this genus, widely separated in their
distribution. The European species, M. vulgaris, the “Hake,” is found
on both sides of the Atlantic, and grows to a length of four feet. It is
caught in great numbers, and preserved as “Stock-fish.” The second
species M. gayi, is common in the Straits of Magelhæn and on the
coast of Chili, less so in New Zealand.
The vertebral column of this genus shows a singular modification
of the apophyses. The neural spines of all the abdominal vertebræ
are extremely strong, dilated, wedged into one another. The
parapophyses of the third to sixth vertebræ are slender, styliform,
whilst those of all the following abdominal vertebræ are very long
and broad, convex on the upper and concave on the lower surface;
the two or three anterior pairs are, as it were, inflated. The whole
forms a strong roof for the air-bladder, reminding us of a similar
structure in Kurtus.
Pseudophycis.—Body of moderate length, covered with rather
small scales. A separate caudal, two dorsals, and one anal; ventral
fins very narrow and styliform, but composed of several rays. Jaws
with a band of small teeth; vomer and palatines toothless. Chin with a
barbel.
Two species, of which Ps. bachus is common on the coast of
New Zealand.
Allied genera are Lotella, Physiculus, Uraleptus, and
Læmonema, from moderate depths, obtained chiefly off Madeira and
the Southern Temperate Zone.